{ "metadata": { "ParamSize": 455, "ParamBytes": 1564948480.0, "BitsPerParam": 4.498393057862795 }, "records": [ { "dataPath": "params_shard_0.bin", "format": "raw-shard", "nbytes": 65536000, "records": [ { "name": "lm_head.linear.q_weight", "shape": [ 51200, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 65536000, "byteOffset": 0 } ], "md5sum": "fc96bf912b02dea817c4d490b44ae917" }, { "dataPath": "params_shard_1.bin", "format": "raw-shard", "nbytes": 23091200, "records": [ { "name": "lm_head.linear.bias", "shape": [ 51200 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 102400, "byteOffset": 0 }, { "name": "lm_head.linear.q_scale", "shape": [ 51200, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 8192000, "byteOffset": 102400 }, { "name": "lm_head.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8294400 }, { "name": "lm_head.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 8299520 }, { "name": "transformer.h.30.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 8304640 }, { "name": "transformer.h.30.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 8320000 }, { "name": "transformer.h.30.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 18150400 }, { "name": "transformer.h.30.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 19379200 }, { "name": "transformer.h.30.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 19384320 }, { "name": "transformer.h.30.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 22661120 }, { "name": "transformer.h.30.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 23070720 } ], "md5sum": "8b59a9368a20b6e4bb74867780b5be55" }, { "dataPath": "params_shard_2.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.30.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.30.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.30.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.30.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.30.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.31.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.31.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.31.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "e45238554b8bdabb9c477cf558e7c9a0" }, { "dataPath": "params_shard_3.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.31.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.31.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.31.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.31.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.31.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.31.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.31.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.31.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.31.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "357d6f6515576f4b712862cbe6348b74" }, { "dataPath": "params_shard_4.bin", "format": "raw-shard", "nbytes": 65536000, "records": [ { "name": "transformer.embd.q_weight", "shape": [ 51200, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 65536000, "byteOffset": 0 } ], "md5sum": "700bb689465499a6f21bcba48b1a85b8" }, { "dataPath": "params_shard_5.bin", "format": "raw-shard", "nbytes": 32793600, "records": [ { "name": "transformer.h.31.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.31.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.embd.q_scale", "shape": [ 51200, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 8192000, "byteOffset": 14745600 }, { "name": "transformer.h.0.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 22937600 }, { "name": "transformer.h.0.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 22942720 }, { "name": "transformer.h.0.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 22947840 }, { "name": "transformer.h.0.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 22963200 } ], "md5sum": "da99992a7fb090cd095545d901340e2f" }, { "dataPath": "params_shard_6.bin", "format": "raw-shard", "nbytes": 32798720, "records": [ { "name": "transformer.h.0.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 0 }, { "name": "transformer.h.0.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 1228800 }, { "name": "transformer.h.0.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 1233920 }, { "name": "transformer.h.0.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 4510720 }, { "name": "transformer.h.0.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 4920320 }, { "name": "transformer.h.0.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 4940800 }, { "name": "transformer.h.0.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 18048000 }, { "name": "transformer.h.0.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 19686400 }, { "name": "transformer.h.0.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 19691520 } ], "md5sum": "d72706d12f09ad36359b3c5231312060" }, { "dataPath": "params_shard_7.bin", "format": "raw-shard", "nbytes": 31185920, "records": [ { "name": "transformer.h.0.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 0 }, { "name": "transformer.h.1.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 1638400 }, { "name": "transformer.h.1.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 1643520 }, { "name": "transformer.h.1.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 1648640 }, { "name": "transformer.h.1.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 1664000 }, { "name": "transformer.h.1.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 11494400 }, { "name": "transformer.h.1.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 12723200 }, { "name": "transformer.h.1.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 12728320 }, { "name": "transformer.h.1.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 16005120 }, { "name": "transformer.h.1.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 16414720 }, { "name": "transformer.h.1.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 16435200 }, { "name": "transformer.h.1.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 29542400 }, { "name": "transformer.h.1.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 31180800 } ], "md5sum": "1f46f8e7d6359cdf20506557591013c6" }, { "dataPath": "params_shard_8.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.1.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.1.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.10.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.10.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.10.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.10.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.10.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.10.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.10.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.10.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.10.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "6cb5c0373220622729e3f28ded14cbe4" }, { "dataPath": "params_shard_9.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.10.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.10.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.10.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.10.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.10.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.11.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.11.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.11.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "66ea15aa5187bd8cad2c265f9beb1148" }, { "dataPath": "params_shard_10.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.11.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.11.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.11.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.11.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.11.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.11.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.11.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.11.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.11.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "384a313abbba3d1d4df83c1356782da3" }, { "dataPath": "params_shard_11.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.11.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.11.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.12.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.12.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.12.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.12.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.12.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.12.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.12.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.12.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.12.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "ee3469d024577b35e9b63f77df6153c0" }, { "dataPath": "params_shard_12.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.12.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.12.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.12.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.12.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.12.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.13.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.13.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.13.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "d8d4b77d2a3b221c3979c000bcdb1554" }, { "dataPath": "params_shard_13.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.13.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.13.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.13.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.13.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.13.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.13.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.13.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.13.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.13.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "d2e650508b0609afc91b57f97c9497c9" }, { "dataPath": "params_shard_14.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.13.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.13.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.14.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.14.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.14.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.14.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.14.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.14.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.14.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.14.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.14.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "a64d6d37cba995908399bf283b876222" }, { "dataPath": "params_shard_15.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.14.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.14.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.14.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.14.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.14.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.15.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.15.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.15.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "fd449b60007db3ca615988793c680512" }, { "dataPath": "params_shard_16.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.15.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.15.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.15.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.15.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.15.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.15.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.15.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.15.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.15.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "47bcf34bbe9be12be1b098a4ecd623d6" }, { "dataPath": "params_shard_17.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.15.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.15.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.16.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.16.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.16.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.16.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.16.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.16.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.16.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.16.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.16.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "faad14fb1a78d2de77ec9a14e4416eb9" }, { "dataPath": "params_shard_18.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.16.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.16.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.16.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.16.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.16.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.17.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.17.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.17.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "9e957f99f15bda9feeafcae5255f6598" }, { "dataPath": "params_shard_19.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.17.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.17.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.17.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.17.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.17.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.17.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.17.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.17.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.17.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "41eda10b405acc35e0407d11eaa08317" }, { "dataPath": "params_shard_20.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.17.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.17.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.18.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.18.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.18.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.18.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.18.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.18.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.18.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.18.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.18.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "dc594d3e8a71d829b441e2cdf56342e0" }, { "dataPath": "params_shard_21.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.18.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.18.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.18.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.18.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.18.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.19.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.19.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.19.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "5a24581ab0339693272a5cd4c9e1e2a2" }, { "dataPath": "params_shard_22.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.19.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.19.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.19.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.19.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.19.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.19.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.19.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.19.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.19.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "db6c8e9e8369958d4f69b2358ac01639" }, { "dataPath": "params_shard_23.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.19.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.19.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.2.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.2.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.2.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.2.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.2.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.2.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.2.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.2.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.2.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "07f4a7c1726e7448ec81457f59c83a77" }, { "dataPath": "params_shard_24.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.2.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.2.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.2.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.2.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.2.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.20.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.20.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.20.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "f09ee8a67c657e4905bb13aca130170f" }, { "dataPath": "params_shard_25.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.20.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.20.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.20.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.20.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.20.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.20.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.20.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.20.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.20.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "966aceffffdd92bed358ef92cad1f10d" }, { "dataPath": "params_shard_26.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.20.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.20.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.21.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.21.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.21.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.21.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.21.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.21.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.21.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.21.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.21.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "6f86202d5f8cb7baccbd07a21666bdd9" }, { "dataPath": "params_shard_27.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.21.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.21.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.21.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.21.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.21.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.22.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.22.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.22.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "d929cfa96d4713a22db165cc77a22cdf" }, { "dataPath": "params_shard_28.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.22.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.22.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.22.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.22.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.22.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.22.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.22.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.22.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.22.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "627b81a1e070893d011cd7b831deff55" }, { "dataPath": "params_shard_29.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.22.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.22.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.23.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.23.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.23.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.23.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.23.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.23.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.23.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.23.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.23.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "c3bc1e5da7caefdaa6d597af99dfdd27" }, { "dataPath": "params_shard_30.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.23.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.23.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.23.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.23.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.23.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.24.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.24.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.24.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "f6195ba3b64373cf7683a0df1cf3932a" }, { "dataPath": "params_shard_31.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.24.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.24.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.24.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.24.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.24.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.24.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.24.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.24.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.24.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "e015c009e49cef074949f2d02c0d98af" }, { "dataPath": "params_shard_32.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.24.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.24.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.25.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.25.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.25.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.25.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.25.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.25.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.25.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.25.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.25.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "b24467883e345aa378fa92018fc10d88" }, { "dataPath": "params_shard_33.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.25.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.25.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.25.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.25.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.25.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.26.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.26.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.26.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "4b965c6d31c55e9c48a0e64d9ec5f810" }, { "dataPath": "params_shard_34.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.26.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.26.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.26.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.26.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.26.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.26.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.26.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.26.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.26.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "819759e5dcd862cdb7955256df76fc7d" }, { "dataPath": "params_shard_35.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.26.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.26.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.27.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.27.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.27.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.27.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.27.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.27.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.27.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.27.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.27.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "efe752d023761489e760ae4acc86f5f8" }, { "dataPath": "params_shard_36.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.27.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.27.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.27.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.27.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.27.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.28.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.28.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.28.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "0b1dd211566511b9d95ec522043df399" }, { "dataPath": "params_shard_37.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.28.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.28.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.28.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.28.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.28.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.28.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.28.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.28.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.28.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "436251f89e9cf25f0ab7d15e6fd7c48b" }, { "dataPath": "params_shard_38.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.28.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.28.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.29.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.29.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.29.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.29.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.29.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.29.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.29.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.29.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.29.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "210ccfd31d8f5a50fe625f8f8c0ef73c" }, { "dataPath": "params_shard_39.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.29.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.29.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.29.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.29.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.29.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.3.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.3.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.3.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "64e2e87945122b12725afb9440e0c32d" }, { "dataPath": "params_shard_40.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.3.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.3.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.3.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.3.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.3.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.3.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.3.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.3.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.3.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "84f8cb70a71d685a248b1bc4fed214dc" }, { "dataPath": "params_shard_41.bin", "format": "raw-shard", "nbytes": 29552640, "records": [ { "name": "transformer.h.3.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.3.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.30.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.30.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.4.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14755840 }, { "name": "transformer.h.4.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14760960 }, { "name": "transformer.h.4.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14766080 }, { "name": "transformer.h.4.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14781440 }, { "name": "transformer.h.4.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24611840 }, { "name": "transformer.h.4.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25840640 }, { "name": "transformer.h.4.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25845760 }, { "name": "transformer.h.4.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29122560 }, { "name": "transformer.h.4.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29532160 } ], "md5sum": "e0d700566f0c46311025eb0e642d7923" }, { "dataPath": "params_shard_42.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.4.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.4.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.4.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.4.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.4.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.5.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.5.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.5.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "c6a13762fbe6ab726e73df45372998be" }, { "dataPath": "params_shard_43.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.5.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.5.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.5.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.5.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.5.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.5.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.5.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.5.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.5.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "37c38fad1c306740f8a38e3dbe0d6815" }, { "dataPath": "params_shard_44.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.5.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.5.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.6.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.6.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.6.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.6.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.6.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.6.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.6.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.6.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.6.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "8f4aef222dc955edddd85d2b9683524f" }, { "dataPath": "params_shard_45.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.6.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.6.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.6.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.6.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.6.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.7.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.7.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.7.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "9653fdc1d105c86dd5dc66451d1a0f70" }, { "dataPath": "params_shard_46.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.7.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.7.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.7.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.7.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.7.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.7.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.7.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.7.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.7.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "dde575b5c178627f131a531f785d6555" }, { "dataPath": "params_shard_47.bin", "format": "raw-shard", "nbytes": 29542400, "records": [ { "name": "transformer.h.7.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.7.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.8.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.8.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14750720 }, { "name": "transformer.h.8.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 14755840 }, { "name": "transformer.h.8.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 14771200 }, { "name": "transformer.h.8.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 24601600 }, { "name": "transformer.h.8.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 25830400 }, { "name": "transformer.h.8.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 25835520 }, { "name": "transformer.h.8.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 29112320 }, { "name": "transformer.h.8.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 29521920 } ], "md5sum": "705173dbb9e7716a01daa77f28551610" }, { "dataPath": "params_shard_48.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.8.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.8.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 }, { "name": "transformer.h.8.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 14745600 }, { "name": "transformer.h.8.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14750720 }, { "name": "transformer.h.8.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27857920 }, { "name": "transformer.h.9.ln.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29496320 }, { "name": "transformer.h.9.ln.weight", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29501440 }, { "name": "transformer.h.9.mixer.Wqkv.bias", "shape": [ 7680 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 15360, "byteOffset": 29506560 } ], "md5sum": "7a15bb6e0575a90b07d184eaee32451c" }, { "dataPath": "params_shard_49.bin", "format": "raw-shard", "nbytes": 29521920, "records": [ { "name": "transformer.h.9.mixer.Wqkv.q_weight", "shape": [ 7680, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 9830400, "byteOffset": 0 }, { "name": "transformer.h.9.mixer.Wqkv.q_scale", "shape": [ 7680, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1228800, "byteOffset": 9830400 }, { "name": "transformer.h.9.mixer.out_proj.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 11059200 }, { "name": "transformer.h.9.mixer.out_proj.q_weight", "shape": [ 2560, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 3276800, "byteOffset": 11064320 }, { "name": "transformer.h.9.mixer.out_proj.q_scale", "shape": [ 2560, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 409600, "byteOffset": 14341120 }, { "name": "transformer.h.9.mlp.fc1.bias", "shape": [ 10240 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 20480, "byteOffset": 14750720 }, { "name": "transformer.h.9.mlp.fc1.q_weight", "shape": [ 10240, 320 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 14771200 }, { "name": "transformer.h.9.mlp.fc1.q_scale", "shape": [ 10240, 80 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 27878400 }, { "name": "transformer.h.9.mlp.fc2.bias", "shape": [ 2560 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 5120, "byteOffset": 29516800 } ], "md5sum": "03faccebc6dec33f5e75987a687c2736" }, { "dataPath": "params_shard_50.bin", "format": "raw-shard", "nbytes": 14745600, "records": [ { "name": "transformer.h.9.mlp.fc2.q_weight", "shape": [ 2560, 1280 ], "dtype": "uint32", "format": "f32-to-bf16", "nbytes": 13107200, "byteOffset": 0 }, { "name": "transformer.h.9.mlp.fc2.q_scale", "shape": [ 2560, 320 ], "dtype": "float16", "format": "f32-to-bf16", "nbytes": 1638400, "byteOffset": 13107200 } ], "md5sum": "64dac7bcf4ca26a9c76ed7f5193879eb" } ] }